sglang
https://github.com/sgl-project/sglang
Python
SGLang is a high-performance serving framework for large language models and multimodal models.
Triage Issues!
When you volunteer to triage issues, you'll receive an email each day with a link to an open issue that needs help in this project. You'll also receive instructions on how to triage issues.
Triage Docs!
Receive a documented method or class from your favorite GitHub repos in your inbox every day. If you're really pro, receive undocumented methods or classes and supercharge your commit history.
Python not yet supported1 Subscribers
Add a CodeTriage badge to sglang
Help out
- Issues
- [Feat] serving: add configurable --max-payload-size for HTTP bodies
- [Perf] scheduler: Add metirc scheduler_iterations_total
- [Feat] LoRA: revamp LoRADrainer for better performance
- [Fix] benchmark: Count multi-turn input tokens in serving metrics
- [Feat] mem_cache: marconi eviction policy for the mamba radix cache
- [Fix] Qwen3: Pass quant_config to Qwen3Attention
- [Feature] Add XPress speculative decoding method
- [Feat] disaggregation: Support Qwen3.8-Flash-Next CPU-buffer PD
- [Fix] api: Enforce integer ranges in generated Python types
- [Bug] Inkling: fused attention prologue stores CUDA-graph padding rows into KV slot 0 → NaN logits (`probability tensor contains either inf, nan or element < 0` with `SGLANG_OPT_USE_GUMBEL_SAMPLE=0`)
- Docs
- Python not yet supported